hi.. Newbie here
can anyone tell me if they have software available to access the Wiimotes camera directly.. or is it not possible.. im thinking it is because if you setup your wiimote on the console itself you can see what the camera sees.. if this is posible you should be able to get more than 4 blobs recognised shouldnt u ?
The Wii Remote isn’t programmed (and thus, doesn’t know how) to throw the raw IR camera feed off on the bluetooth. One thing you need to understand about the Wii and Wii Remote is that while they do use Bluetooth to communicate with each other, the Wii’s implementation of the bluetooth stack is very bare-bones and specialized for their application. Not that it’s easy to do remote camera feeds over bluetooth anyway - bluetooth wasn’t originally designed to replace (or even implement) wireless camera protocols.
That’s the reason why the Wii remote only supports tracking of no more than 4 IR blobs - that’s what the bluetooth interface provides. It abstracts the camera away, and only provides the X/Y/Intensity values of 4 blobs through the bluetooth interface. The Wii Remote sends the blob data over the bluetooth, the Wii responds by translating the blob data into the cursor’s position. The Wii doesn’t even realize that the data came from an IR camera, it just needs the blobs.
The only way to “access the camera directly” would be to destroy the Wii remote by physically removing the camera assembly, and programming a micro-controller to talk to it. It would take a lot of research on the specific camera parts (specs, how to program it, etc), and assembly programming experience for the micro-controller you decide upon.
Years ago, I took an Assembly programming language class - one of the hardest classes I ever took. Our final project was to program an Atmel micro-controller to talk to an IR sensor capable of range finding (useful for a robot that needs to see how far away it is from objects in the vicinity, among other things). The data the sensor provided was fairly one-dimensional, so the process was fairly straight-forward. Reading a Camera’s CCD isn’t so straight-forward, since there’s exponentially more data to have to go through. Ideally, you’d need a pretty powerful GPU capable of reading over the 2-dimensional image the camera would provide, since a GPU is a little more specialized than a CPU when it comes to parallel operations - a CPU would have to apply a function to every element, one at a time, in the grid that a CCD provides - a GPU would be able to blast the same function at a grid of data all at once.